Compare / Qwen3.5-122B-A10B vs Qwen3.5-27B
Qwen3.5-122B-A10B vs Qwen3.5-27B
Which is cheaper, which is more capable, and which does better at each kind of work, from 19 results on 4 sources that measured both models.
Of the 19 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.
Results as of 13 October 2026 (catalogue release 2026-10-13-2e2420e3c5cc).
Qwen3.5-122B-A10B or Qwen3.5-27B? The short answer
Pick Qwen3.5-27B: it scores higher (Intelligence Index 22.9 against 17.7) at about the same price.
- They cost about the same: $0.72 and $0.74 per million tokens (three input to one output).
- Qwen3.5-27B scores higher on the Artificial Analysis Intelligence Index: 22.9 against 17.7.
- On third-party benchmarks, on the results both models have: Qwen3.5-122B-A10B does better for sticking to the facts; Qwen3.5-27B for reasoning and knowledge and writing.
Head to head
Every fact side by side
| Qwen3.5-122B-A10B | Qwen3.5-27B | |
|---|---|---|
| Price per million tokens | $0.26 in · $2.08 out | $0.27 in · $2.16 out |
| Blended (3 in : 1 out) | $0.72 | $0.74 |
| Cheapest host, blended | $0.72 (Alibaba) | $0.54 (Alibaba) |
| Intelligence Index | 17.7 | 22.9 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | 147 | – |
| Context window | 262,144 tokens | 262,144 tokens |
| Released | 24 Feb 2026 | 24 Feb 2026 |
| Weights | Open | Open |
| Developer based in | China | China |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.
Every shared result
Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.
| Measure | Qwen3.5-122B-A10B | Qwen3.5-27B |
|---|---|---|
| Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results | ||
| Overall · TextArena rating, higher is better 117 · 129 of 387 measured | 1,416 | 1,409 |
| Overall · Text factualityArena rating, higher is better 103 · 112 of 163 measured | 1,428 | 1,420 |
| Business, management and finance · TextArena rating, higher is better 105 · 128 of 380 measured | 1,422 | 1,408 |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 114 · 86 of 243 measured | 17.7no reasoning | 22.9 |
| GPQA Diamond% of questions, higher is better 75 · 74 of 231 measured | 85.7% | 85.8% |
| Humanity's Last Exam% of questions, higher is better 92 · 95 of 242 measured | 25.2% | 23.9% |
| UGI Leaderboard reported by UGI Leaderboard · 3 results | ||
| Writing score · UGIscore out of 100, higher is better 112 · 99 of 230 measured | 39.5no reasoning | 42.4think-prefill reasoning |
| Requested-length error · UGI% off the requested word count, lower is better 33 · 49 of 230 measured | 9.0%thinking reasoning | 11.0%think-prefill reasoning |
| Style adherence · UGIscore from 0 to 1, higher is better 83 · 147 of 230 measured | 0.36no reasoning | 0.33no reasoning |
| Vectara Hallucination Leaderboard reported by Vectara Hallucination Leaderboard · 2 results | ||
| Hallucination rate · Vectara% of summaries, lower is better 72 · 80 of 104 measured | 11.2% | 12.1% |
| Answer rate · Vectara% of documents, higher is better 31 · 31 of 104 measured | 99.8% | 99.8% |
Show all 19 results (8 more)
| Measure | Qwen3.5-122B-A10B | Qwen3.5-27B |
|---|---|---|
| Arena (formerly LMArena) reported by Arena (formerly LMArena) · 4 results | ||
| Creative writing · TextArena rating, higher is better 136 · 145 of 385 measured | 1,366 | 1,358 |
| Expert prompts · TextArena rating, higher is better 111 · 112 of 338 measured | 1,442 | 1,441 |
| Instruction following · TextArena rating, higher is better 115 · 122 of 387 measured | 1,406 | 1,401 |
| Writing, literature and language · TextArena rating, higher is better 125 · 135 of 386 measured | 1,387 | 1,380 |
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| IFBench% of instructions, higher is better 22 · 24 of 195 measured | 75.7% | 75.6% |
| Long-context reasoning (AA-LCR)% of questions, higher is better 72 · 63 of 233 measured | 76.3% | 77.7% |
| Terminal-Bench Hard% of tasks, higher is better 65 · 61 of 190 measured | 31.1% | 32.6% |
| τ²-bench telecom% of tasks, higher is better 30 · 28 of 193 measured | 93.6% | 93.9% |
Quick answers
Is Qwen3.5-122B-A10B better than Qwen3.5-27B?
Pick Qwen3.5-27B: it scores higher (Intelligence Index 22.9 against 17.7) at about the same price.
Which is cheaper, Qwen3.5-122B-A10B or Qwen3.5-27B?
They cost about the same: $0.72 and $0.74 per million tokens (three input to one output).
Which is better for agents and tool use, Qwen3.5-122B-A10B or Qwen3.5-27B?
Qwen3.5-122B-A10B ranks 95 of 164 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 145 for Qwen3.5-27B.
Which is better for reasoning and knowledge, Qwen3.5-122B-A10B or Qwen3.5-27B?
Qwen3.5-27B ranks 82 of 198 models on sale for hard questions across science, maths and general knowledge, against 87 for Qwen3.5-122B-A10B.
Which is better for writing, Qwen3.5-122B-A10B or Qwen3.5-27B?
Qwen3.5-27B ranks 98 of 188 models on sale for what people prefer in blind comparisons, and judged writing quality, against 102 for Qwen3.5-122B-A10B.
Which is better for sticking to the facts, Qwen3.5-122B-A10B or Qwen3.5-27B?
Qwen3.5-122B-A10B ranks 108 of 152 models on sale for summarising without inventing things, and factual answers, against 121 for Qwen3.5-27B.
Run Qwen3.5-122B-A10B and Qwen3.5-27B on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.